<p>This research investigates the morphological structure of adjectives in Dogri, a low-resource Indo-Aryan language, using a paradigm-based approach to identify inflectional patterns across gender, number, and case The primary objective is to design a systematic framework that not only captures the intricacies of Dogri adjective morphology but also contributes to the foundational development of a morphological analyzer for the language. A structured dataset of 25,830 adjectives was compiled, and six inflectional paradigms were constructed to map 25,380 unique forms. To assess the framework’s effectiveness, two stemming techniques a hybrid rule-based stemmer and a deep learning-based stemmer were applied and evaluated using precision, recall, and F1-score across vocative, oblique, direct, singular, plural, and gender-based cases. The deep learning model outperformed the hybrid approach, achieving an average accuracy of 90.75% compared to 85.78%. The comparative results underscore the deep learning model’s ability to capture subtle morphological nuances. A focused error analysis reveals challenges in handling irregular inflections, borrowed forms, and compound adjectives. This work not only advances the linguistic modeling of Dogri but also lays a foundation for broader NLP applications such as POS tagging and machine translation for low-resource languages.</p>